CONTEXT

Self-driving cars and traditional cars and increasingly sharing roads today. In such a hybrid world, we have to envision new modes of human-machine communication that can connect people and automobiles in a universal manner.

INSIGHT

IDEA

intentional gestures made by drivers as well as subconscious inferences made by pedestrians and observers constitute an unspoken yet critical component of interactions at urban intersections.

Enabling vehicles to communicate directly with other drivers and pedestrians outside  through gestures allows to establish new feedback loops which can help maintain sustained awareness of how driverless vehicles navigate human environments. 

IMPACT

The proposed concept allowed pedestrians to predict with certainty, the the actions self-driving vehicles were about to make and thereby relieve passengers of communication tasks which they would otherwise have no control of. 

TEAM

Emil Alex

ADVISORS

Sara Dean 

Scott Minneman

Maria Mortati

TOOLS AND METHODOLOGIES

User Surveys / Field Research / Contextual Inquiry  Journey Mapping / Rapid Prototyping / CAD Modeling / User Testing / 

COMMUNICATION MODELS 

FOR DRIVERLESS VEHICLES

GRADUATE THESIS PROJECT

THE CHALLENGE

As driverless vehicles are becoming a reality we are in a phase where they will share the roads with cars that have human beings at the wheel. This phase introduces a new set of challenges in terms of signaling and communication.focussed on the communication channels involving the direct interaction and indirect inferences that drivers and pedestrians make while navigating urban intersections.

THE CONTEXT

SCOPING THE AREAS OF WORK
KEY QUESTIONS
How would an automobile communicate with people outside  in a driverless world?
What are the interactions that currently take place between :
  • Driver and other drivers
  • Driver and pedestrians
  • Pedestrians and the environment
What are the activities that a driver is involved in while driving?
What are the factors that affect a pedestrian while the navigate intersections?
How would an automobile communicate with people outside  in a driverless world?
Drivers were recorded along with corresponding video from the outside of the car as they navigated the urban environment. Factors tracked included:
​​
  • Movement
  • Triggers 
  • Reaction times
  • Gestures
RECORDING DRIVER BEHAVIOR

FEEDBACK

CHANNELS

PLATFORM

AUDIO

VISUAL

HEADLIGHTS

TURN

SIGNALS

BRAKE

LIGHTS

REVERSE

LIGHTS

HORN

INFORMING

BEHIND

INDICATING

TURNS

ALERTING

ONCOMING TRAFFIC

GETTING THE ATTENTION OF VEHICLES AND PEDESTRIANS NEARBY

SIGNALLING DECELERATION

GESTURES

OTHER DRIVERS

PEDESTRIANS

INCHING AHEAD AT INTERSECTIONS

GIVING

PERMISSION

SLOWING DOWN TO ALLOW LANE CHANGES

REQUESTING

PERMISSION

INFERENCES

AUDIO

VISUAL

BEING AWARE OF WHERE OTHER DRIVERS ARE LOOKING

SENSING THE PRESENCE OF ANOTHER CAR BASED ON SOUND

DETECTING THE REACTION TO THE HONKING OF A HORN

1/5
KEY ASPECTS
SUBSTITUTING THE DRIVER'S GLANCE THAT ACKNOWLEDGES THE PEDESTRIAN'S PRESENCE
ANTICIPATORY
NOTIFICATIONS
DIRECT
GESTURES
LOCATION AWARENESS
SIGNALING STARTS, STOPS AND TURNS
CONVEYING INFORMATION TO PEDESTRIANS AND OTHER DRIVERS

INSIGHTS

SYNTHESISING THE RESEARCH DATA
 
Pedestrians interact extensively with drivers and their vehicles at urban intersections. These include non-verbal communication along with a subconscious awareness of their environment

Taking all these factors into account, physical forms were envisioned that helped communicate signals with clarity and ease.

From the forms that were generated, a monolithic volume was selected as the most appropriate form for the purpose and developed further using CAD software. This allowed for the fine tuning of details that would enable greater efficiency.

The final CAD model highlights the physical features of an automotive form that is tailored for the purpose of conveying clear and universal messages to pedestrians outside.

1/4

1/4

The final version of the model is presently being developed in a Virtual Reality environment using Autodesk VRED in order to facilitate further user testing

EMIL ALEX 2018